This invention relates to a fingerprint image processing system for use in processing a fingerprint image prior to identification of a fingerprint.
In general, a fingerprint is widely used for identifying a person due to uniqueness, idiosyncrasy, and unchangeability of the fingerprint. Recently, such identification of the fingerprint is automatically carried out by the use of a pattern recognition technique in an electronic computer. For this purpose, ridge endings, bifurcations are extracted from a fingerprint image together with ridges and collated with those of registered or memorized fingerprints. The ridge endings, the bifurcations may be called minutiae hereinunder.
Herein, it is to be noted that a fingerprint image left at a criminal or an accidental spot or place is often extracted from an object which might be decorated itself and/or which is not flat. Under the circumstances, the fingerprint image is not clear on the minutiae and the ridges but is very often obscure. In other words, such a fingerprint image is very often contaminated by such as superfluous lines and dots which might be wrongly recognized as ridges and minutiae and adversely affected by any other defects, such as undesirable deletion of true minutiae and undesirable separation of ridges. Therefore, it is difficult to identify a person by the use of such an unclear fingerprint image.
Taking the above into consideration, the unclear fingerprint image is traced on tracing paper or the like along the ridges to obtain a traced fingerprint image. Subsequently, the traced fingerprint image is memorized into a memory and is displayed on a display device, such as a CRT. Thereafter, minutiae are extracted from the traced fingerprint image and stored in a memory in the form of minutia data signals.
However, a trace operation of the unclear fingerprint image imposes a heavy load on an operator and is time-consuming. In addition, it is not easy to add and delete minutiae to and from the traced fingerprint image and to modify them.